1
0
Fork 0
mirror of https://github.com/moby/moby.git synced 2022-11-09 12:21:53 -05:00
moby--moby/utils.go
Dominik Honnef 22f1cc955d replace unreachable returns with panics
Not only is this a more common idiom, it'll make finding bugs easier,
and it'll make porting to Go 1.1 easier.

Go 1.1 will not require the final return or panic because it has a
notion of terminating statements.
2013-04-03 11:18:23 +02:00

343 lines
8.3 KiB
Go

package docker
import (
"bytes"
"errors"
"fmt"
"github.com/dotcloud/docker/rcli"
"index/suffixarray"
"io"
"io/ioutil"
"net/http"
"os"
"os/exec"
"path/filepath"
"runtime"
"strings"
"sync"
"time"
)
// Go is a basic promise implementation: it wraps calls a function in a goroutine,
// and returns a channel which will later return the function's return value.
func Go(f func() error) chan error {
ch := make(chan error)
go func() {
ch <- f()
}()
return ch
}
// Request a given URL and return an io.Reader
func Download(url string, stderr io.Writer) (*http.Response, error) {
var resp *http.Response
var err error = nil
if resp, err = http.Get(url); err != nil {
return nil, err
}
if resp.StatusCode >= 400 {
return nil, errors.New("Got HTTP status code >= 400: " + resp.Status)
}
return resp, nil
}
// Debug function, if the debug flag is set, then display. Do nothing otherwise
// If Docker is in damon mode, also send the debug info on the socket
func Debugf(format string, a ...interface{}) {
if os.Getenv("DEBUG") != "" {
// Retrieve the stack infos
_, file, line, ok := runtime.Caller(1)
if !ok {
file = "<unknown>"
line = -1
} else {
file = file[strings.LastIndex(file, "/")+1:]
}
fmt.Fprintf(os.Stderr, fmt.Sprintf("[debug] %s:%d %s\n", file, line, format), a...)
if rcli.CLIENT_SOCKET != nil {
fmt.Fprintf(rcli.CLIENT_SOCKET, fmt.Sprintf("[debug] %s:%d %s\n", file, line, format), a...)
}
}
}
// Reader with progress bar
type progressReader struct {
reader io.ReadCloser // Stream to read from
output io.Writer // Where to send progress bar to
readTotal int // Expected stream length (bytes)
readProgress int // How much has been read so far (bytes)
lastUpdate int // How many bytes read at least update
}
func (r *progressReader) Read(p []byte) (n int, err error) {
read, err := io.ReadCloser(r.reader).Read(p)
r.readProgress += read
// Only update progress for every 1% read
updateEvery := int(0.01 * float64(r.readTotal))
if r.readProgress-r.lastUpdate > updateEvery || r.readProgress == r.readTotal {
fmt.Fprintf(r.output, "%d/%d (%.0f%%)\r",
r.readProgress,
r.readTotal,
float64(r.readProgress)/float64(r.readTotal)*100)
r.lastUpdate = r.readProgress
}
// Send newline when complete
if err == io.EOF {
fmt.Fprintf(r.output, "\n")
}
return read, err
}
func (r *progressReader) Close() error {
return io.ReadCloser(r.reader).Close()
}
func ProgressReader(r io.ReadCloser, size int, output io.Writer) *progressReader {
return &progressReader{r, output, size, 0, 0}
}
// HumanDuration returns a human-readable approximation of a duration
// (eg. "About a minute", "4 hours ago", etc.)
func HumanDuration(d time.Duration) string {
if seconds := int(d.Seconds()); seconds < 1 {
return "Less than a second"
} else if seconds < 60 {
return fmt.Sprintf("%d seconds", seconds)
} else if minutes := int(d.Minutes()); minutes == 1 {
return "About a minute"
} else if minutes < 60 {
return fmt.Sprintf("%d minutes", minutes)
} else if hours := int(d.Hours()); hours == 1 {
return "About an hour"
} else if hours < 48 {
return fmt.Sprintf("%d hours", hours)
} else if hours < 24*7*2 {
return fmt.Sprintf("%d days", hours/24)
} else if hours < 24*30*3 {
return fmt.Sprintf("%d weeks", hours/24/7)
} else if hours < 24*365*2 {
return fmt.Sprintf("%d months", hours/24/30)
}
return fmt.Sprintf("%d years", d.Hours()/24/365)
}
func Trunc(s string, maxlen int) string {
if len(s) <= maxlen {
return s
}
return s[:maxlen]
}
// Figure out the absolute path of our own binary
func SelfPath() string {
path, err := exec.LookPath(os.Args[0])
if err != nil {
panic(err)
}
path, err = filepath.Abs(path)
if err != nil {
panic(err)
}
return path
}
type nopWriteCloser struct {
io.Writer
}
func (w *nopWriteCloser) Close() error { return nil }
func NopWriteCloser(w io.Writer) io.WriteCloser {
return &nopWriteCloser{w}
}
type bufReader struct {
buf *bytes.Buffer
reader io.Reader
err error
l sync.Mutex
wait sync.Cond
}
func newBufReader(r io.Reader) *bufReader {
reader := &bufReader{
buf: &bytes.Buffer{},
reader: r,
}
reader.wait.L = &reader.l
go reader.drain()
return reader
}
func (r *bufReader) drain() {
buf := make([]byte, 1024)
for {
n, err := r.reader.Read(buf)
r.l.Lock()
if err != nil {
r.err = err
} else {
r.buf.Write(buf[0:n])
}
r.wait.Signal()
r.l.Unlock()
if err != nil {
break
}
}
}
func (r *bufReader) Read(p []byte) (n int, err error) {
r.l.Lock()
defer r.l.Unlock()
for {
n, err = r.buf.Read(p)
if n > 0 {
return n, err
}
if r.err != nil {
return 0, r.err
}
r.wait.Wait()
}
panic("unreachable")
}
func (r *bufReader) Close() error {
closer, ok := r.reader.(io.ReadCloser)
if !ok {
return nil
}
return closer.Close()
}
type writeBroadcaster struct {
mu sync.Mutex
writers map[io.WriteCloser]struct{}
}
func (w *writeBroadcaster) AddWriter(writer io.WriteCloser) {
w.mu.Lock()
w.writers[writer] = struct{}{}
w.mu.Unlock()
}
// FIXME: Is that function used?
// FIXME: This relies on the concrete writer type used having equality operator
func (w *writeBroadcaster) RemoveWriter(writer io.WriteCloser) {
w.mu.Lock()
delete(w.writers, writer)
w.mu.Unlock()
}
func (w *writeBroadcaster) Write(p []byte) (n int, err error) {
w.mu.Lock()
defer w.mu.Unlock()
for writer := range w.writers {
if n, err := writer.Write(p); err != nil || n != len(p) {
// On error, evict the writer
delete(w.writers, writer)
}
}
return len(p), nil
}
func (w *writeBroadcaster) CloseWriters() error {
w.mu.Lock()
defer w.mu.Unlock()
for writer := range w.writers {
writer.Close()
}
w.writers = make(map[io.WriteCloser]struct{})
return nil
}
func newWriteBroadcaster() *writeBroadcaster {
return &writeBroadcaster{writers: make(map[io.WriteCloser]struct{})}
}
func getTotalUsedFds() int {
if fds, err := ioutil.ReadDir(fmt.Sprintf("/proc/%d/fd", os.Getpid())); err != nil {
Debugf("Error opening /proc/%d/fd: %s", os.Getpid(), err)
} else {
return len(fds)
}
return -1
}
// TruncIndex allows the retrieval of string identifiers by any of their unique prefixes.
// This is used to retrieve image and container IDs by more convenient shorthand prefixes.
type TruncIndex struct {
index *suffixarray.Index
ids map[string]bool
bytes []byte
}
func NewTruncIndex() *TruncIndex {
return &TruncIndex{
index: suffixarray.New([]byte{' '}),
ids: make(map[string]bool),
bytes: []byte{' '},
}
}
func (idx *TruncIndex) Add(id string) error {
if strings.Contains(id, " ") {
return fmt.Errorf("Illegal character: ' '")
}
if _, exists := idx.ids[id]; exists {
return fmt.Errorf("Id already exists: %s", id)
}
idx.ids[id] = true
idx.bytes = append(idx.bytes, []byte(id+" ")...)
idx.index = suffixarray.New(idx.bytes)
return nil
}
func (idx *TruncIndex) Delete(id string) error {
if _, exists := idx.ids[id]; !exists {
return fmt.Errorf("No such id: %s", id)
}
before, after, err := idx.lookup(id)
if err != nil {
return err
}
delete(idx.ids, id)
idx.bytes = append(idx.bytes[:before], idx.bytes[after:]...)
idx.index = suffixarray.New(idx.bytes)
return nil
}
func (idx *TruncIndex) lookup(s string) (int, int, error) {
offsets := idx.index.Lookup([]byte(" "+s), -1)
//log.Printf("lookup(%s): %v (index bytes: '%s')\n", s, offsets, idx.index.Bytes())
if offsets == nil || len(offsets) == 0 || len(offsets) > 1 {
return -1, -1, fmt.Errorf("No such id: %s", s)
}
offsetBefore := offsets[0] + 1
offsetAfter := offsetBefore + strings.Index(string(idx.bytes[offsetBefore:]), " ")
return offsetBefore, offsetAfter, nil
}
func (idx *TruncIndex) Get(s string) (string, error) {
before, after, err := idx.lookup(s)
//log.Printf("Get(%s) bytes=|%s| before=|%d| after=|%d|\n", s, idx.bytes, before, after)
if err != nil {
return "", err
}
return string(idx.bytes[before:after]), err
}
// TruncateId returns a shorthand version of a string identifier for convenience.
// A collision with other shorthands is very unlikely, but possible.
// In case of a collision a lookup with TruncIndex.Get() will fail, and the caller
// will need to use a langer prefix, or the full-length Id.
func TruncateId(id string) string {
shortLen := 12
if len(id) < shortLen {
shortLen = len(id)
}
return id[:shortLen]
}